thermocouples

drewcrew6 · Dec 21, 2002

Okay I'm not sure if I'm going to state this right but here goes.
The accuracy of a thermocouple is "x" does this mean the repeatability of the signal at the same temps or accuracy from one tc to another?

If this doesn't deal with the repeatability then how can that be found?
accuracy chart for tc's
What I'm looking to find out is if the tc sees say a variation of true temp 100 to 110 degrees how accuratly will the same tc "signal" 105.0 degrees .

Thanks
Drewcrew6

rogerhollingsworth · Dec 21, 2002

The accuracy of a thermocouple is dependent on the quality of material that the thermocouple is constructed of. The more pure the material the more accurate the thermocouple will mimic the actual temperature.
There is an acceptable range of material quality that can be used to make up, say, a “J” thermocouple and have it still be called a “J” thermocouple. That range could be X %, I am not sure what it is, so the accuracy and repeatability should fall within that X % for a thermocouple with that same quality of material.
Don’t take this to the bank unless some of the other guys agree, it is just my understanding.

Roger

rsdoran · Dec 21, 2002

drew your link gives the answer, using Rogers J type at 105 degrees the maximum error would be 1.5 degrees, I would assume this is plus or minus which would give approximately a 3% error. From just taking a quick look at the table you provided that 3% is probably the worst case scenario, seems they can be more accurate at higher temps.

BTW thanks for that link, it will help

I just reread your question and now I am not sure we answered the question correctly. A thermocouple should not see a variation between degrees, it should offer a fairly stable value dependent on temp. Using your 105 degree then it should be accurate from 103.5 to 106.5. NOTE: I assumed 1.5 + or -, it may be a flat 1.5 which would mean between 104.25 and 105.75. This part I dont know about but would feel safe with either.

tomneth · Dec 22, 2002

I looked at the chart and it specifically states that is 'absolute' accuracy which I take to mean the maximum error between the thermocouple and a thermometer or other instrument whose accuracy would be the standard. If I remember correctly a thermocouple is not 100% linear, there is some degree of error off a linear line of say 0 - 500 degrees. I believe the different materials have thier maximum error at different temperatures and that is what the chart is showing.

The 'relative' accuracy would mean the repeatability between two thermocouples of the same type measuring the same temperature, even if it was 1 or 2 degrees different than the 'real' temperature.

randylud · Dec 22, 2002

Thermocouples

Drew,
Here is some good info from Omega:
http://www.omega.com/temperature/Z/pdf/z021-032.pdf

It is a pretty long download if you are using dial-up but it is worth the wait. The thermocouple is a non-linear device, plus all the charts are based at 32 deg. F so you know that it is gonna get you close but not perfect. The thermocouple is best applied to systems where extreme accuracy is not required. There are other methods which are more accurate and that includes repeatability. Check out this site and it may help you decide what is right for your app.

drewcrew6 · Dec 22, 2002

I took the accuracy to mean that from one tc to another this is the max difference you will see. From what I'm reading from everyone and on my own it also means the repeatability of the same tc.

What I'm looking to do is monitor room temperature and have that info available on an ab plc.
Looking for a nice clean option with minimal cost and decent accuracy/repeatability.

I may have found my solution got to look at some specs first.

Serial temp module

This guy has a few different items.
His ordering and items page
Now to get the comms between the two if I decide on that setup. That may bring some q's on here but we'll see looks pretty basic to communicate with.

Thanks
Drewcrew6

Tom Jenkins · Dec 23, 2002

Total Accuracy

The accuracy of the thermocouple itself is only part of the total accuracy picture. The wiring, EMI/RFI, and especially the read out or PLC input accuracy all influence the total system accuracy.

You should also ask yourself what accuracy is required by the process. There is no point trying to get to ±0.5% for HVAC applications, for example, but that may be needed for sterilization or for a chemical process control application.

john paley · Dec 23, 2002

Just a point--when mfr's talk of accuracy in terms of %, it usually means % of full scale--not indicated temperature.

okiebob · Dec 23, 2002

Sensor's and the environment

I recently attended a course by Pete Stein and Pat Walters on measurement systems engineering and had my eyes opened in regards to sensors usage and accuracy. One point to consider, HOW does the sensor affect the environment it is to measure. Another dandy the INDIVIDUALITY of the sensor and its leads. Two spools of wire from different lots with an outer sleave. Are the wires coiled (emi) or layed against each other, are the high voltage runs seperate and other GREAT tips. Do a search on Pete and Pat and get an idea of what is happening, pretty interesting stuff just a whole lot to absorb. In a tightly controlled environment/process VERY GOOD to know, in a loosly (sp) controlled area no big deal. But anyhow a very good reference for the future.

estover · Dec 26, 2002

Accuracy and repeatability are not the same.

Accuracy - is the degree of conformity of the output of a measuring instrument to the udeal value of the measured variable...ie highest accuracy is closest dart to the bullseye.

Repeatability - is a measure of the dispersion of the measurements. Standard deviation is also a measure of dispersion....ie good repeatability is when all darts are grouped close together...but that does not require them to be in the bullseye.

If all darts are in the bullseye then it can be considered repeatable and accurate.

Steve Etter · Dec 26, 2002

accuracy vs. repeatability

I don't have anything to offer regarding the tc issue itself, but I do want to point out one thing regarding accuracy compared to repeatability...these two are not the same. I have found from experience that some people are not always aware of the difference.

Here is an example to demonstrate:

Imagine firing a rifle five times at the center of a target.
You hit the target all five times in a nice, tight group.
This tight group, however, is off-center of the target.

In this scenario, we see repeatability (tight grouping) but not accuracy (off-center).

Steve

Steve Etter · Dec 26, 2002

Dang.

Beaten to the punch by mere seconds.

banghead

Steve

drewcrew6 · Dec 26, 2002

The whole problem with the difference between accuracy and repeatablity is that no one states both. Actualy the acuuracy isn't as important as repeatability for what I need to do.

If the sensor tells me that at 105 its 110 then i can offset for that but the offset does nothing if one time it tells me 110 and the next time it tells me 108 then 112, etc, etc.

I'm just need a sensor with good repeatability.

Drewcrew6

mdonner · Dec 28, 2002

Thermocouple repearability

Hi

You may try the following web site:
http://www.apogee-inst.com/irt_spec.htm#IRTS-S

You may find many more using google search using
"Thermocouple repearability"

thermocouples

drewcrew6

Member

rogerhollingsworth

Member

rsdoran

Lifetime Supporting Member

tomneth

Member

randylud

Member

drewcrew6

Member

Tom Jenkins

Lifetime Supporting Member

john paley

Member

okiebob

Member

estover

Member

Steve Etter

Lifetime Supporting Member + Moderator

Steve Etter

Lifetime Supporting Member + Moderator

drewcrew6

Member

mdonner

Member

Similar Topics